Production pipelines for ML in Academia

* Benchmarking?

* Reproducibilty?

*  Establish common base? Difficulties?

* 

Reproducible res  academic literature

* Benchmarking?

Benchmark suite, common datasets, 

* Reproducibilty?

ACM Sigmod

*  Establish common base? Difficulties?

* Repository

Existing tools for reproducibility

* Jupyter notebooks (including GCP Datalab, Google Collaboratory)

* Sigmod is ACM  

*  Establish common base? Difficulties?

*